The New England Journal of Medicine A NOSOCOMIAL OUTBREAK OF FLUOROQUINOLONE-RESISTANT SALMONELLA INFECTION

Transcription

1 A NOSOCOMIAL OUTBREAK OF FLUOROQUINOLONE-RESISTANT SALMONELLA INFECTION SONJA J. OLSEN, PH.D., M.S., EMILIO E. DEBESS, D.V.M., M.P.V.M., TERESA E. MCGIVERN, B.S., NINA MARANO, D.V.M., M.P.H., TOM EBY, B.A., STEVE MAUVAIS, B.A., VIJAY K. BALAN, M.S., GERALD ZIRNSTEIN, PH.D., PAUL R. CIESLAK, M.D., AND FREDERICK J. ANGULO, D.V.M., PH.D. ABSTRACT Background Infection with fluoroquinolone-resistant strains of salmonella is rare, as is nosocomial salmonella infection. We describe the first recognized outbreak of fluoroquinolone-resistant salmonella infection in the United States, which occurred in two nursing homes and one hospital in Oregon. Methods We interviewed medical staff and reviewed patients charts and death certificates. In Nursing Home A we conducted a case control study. Patients were defined as residents of the nursing home from whom fluoroquinolone-resistant Salmonella enterica serotype Schwarzengrund was isolated between February 1996 and December Controls were residents with similar medical conditions whose cultures did not yield salmonella. Results Eleven patients with fluoroquinolone-resistant salmonellosis were identified at two nursing homes. In nine patients a urine culture was positive, in one a stool culture, and in one a wound culture. The index patient had been hospitalized in the Philippines and had probably acquired the infection there. Transmission was probably direct (from patient to patient) or through contact with contaminated surfaces. Treatment with fluoroquinolones during the six months before a culture was obtained was associated with a significant risk of salmonella infection. More fluoroquinolones were used at Nursing Home A than at similar nursing homes in Oregon. The isolates from the outbreak had similar patterns on pulsed-field gel electrophoresis and the same gyra mutations. The isolates from the outbreak were also similar to the only previous isolate of fluoroquinolone-resistant salmonella in the United States, which came from a patient in New York who had been transferred from a hospital in the Philippines. Conclusions We describe a prolonged nosocomial outbreak of infection with fluoroquinolone-resistant S. enterica serotype Schwarzengrund. More such outbreaks are likely in institutional settings, particularly those in which there is heavy use of antimicrobial agents. (N Engl J Med 2001;344: ) Copyright 2001 Massachusetts Medical Society. EACH year an estimated 1.4 million salmonella infections occur in the United States. 1 Most occur as a result of eating contaminated food, 2 particularly foods of animal origin. Person-to-person and nosocomial transmission have become rare in the United States 3 but remain problems in developing countries. 4,5 Antimicrobial agents are not essential for the treatment of most salmonella infections, but they can be lifesaving in cases of severe infection. 6 Fluoroquinolones such as ciprofloxacin are commonly used for adult patients with salmonella infections and for the treatment of acute gastroenteritis. The use of ampicillin, chloramphenicol, and trimethoprim sulfamethoxazole is limited because of increasing antimicrobial resistance to these agents. 7,8 Case reports of fluoroquinolone-resistant salmonella infections are rare; only one case has been reported in the United States Most cases have occurred in persons with underlying illness, particularly persons with human immunodeficiency virus infection and with recurrent salmonella septicemia who have been treated with fluoroquinolones. The one previously reported fluoroquinolone-resistant salmonella infection in the United States was detected in 1995, in a patient who had been hospitalized in the Philippines for a diarrheal illness and who was then transferred to a hospital in New York. A stool sample obtained from the patient in New York yielded Salmonella enterica serotype Schwarzengrund a serotype that is rare in the United States that was resistant to fluoroquinolones. 11 We report the first recognized outbreak of fluoroquinolone-resistant salmonella infection in the United States, also caused by S. enterica serotype Schwarzengrund. This outbreak is notable for two reasons: the salmonella infections were acquired nosocomially, and the outbreak was difficult to control and continued for several years. From the Foodborne and Diarrheal Diseases Branch, Division of Bacterial and Mycotic Diseases, National Center for Infectious Diseases (S.J.O., N.M., G.Z., F.J.A.), and the Epidemic Intelligence Service, Division of Applied Public Health Training, Epidemiology Program Office, (S.J.O.), Centers for Disease Control and Prevention, Atlanta; the Oregon Department of Human Services, Health Division, Portland (E.E.D., T.E.M., S.M., V.K.B., P.R.C.); and the Multnomah County Department of Health, Portland, Oreg. (T.E.). Address reprint requests to Dr. Olsen at the Foodborne and Diarrheal Diseases Branch, Centers for Disease Control and Prevention, 1600 Clifton Rd., Mailstop A-38, Atlanta, GA 30333, or at sco2@ cdc.gov N Engl J Med, Vol. 344, No. 21 May 24,

2 A NOSOCOMIAL OUTBREAK OF FLUOROQUINOLONE-RESISTANT SALMONELLA INFECTION Outbreak METHODS In February 1997, the Oregon Health Division became aware of an outbreak of S. enterica serotype Schwarzengrund infections when three culture-confirmed cases were reported by the health departments of Multnomah and Washington counties. Interviews with the patients revealed that two were current residents and one a past resident of the same nursing home, Nursing Home A (Fig. 1). The index patient had been hospitalized in the Philippines for three months after a severe pontine cerebrovascular accident that occurred during travel and was transferred to a hospital in the United States for long-term care in December Two months later, a culture of a purulent discharge from the site of insertion of a suprapubic urinary catheter yielded S. enterica serotype Schwarzengrund. Stool cultures obtained 12 months later revealed that the patient was still shedding the same organism. The Oregon Health Division inspected Nursing Home A and recommended enhanced infection-control procedures. In March 1997, the Oregon Health Division collected stool cultures from 115 patients and 46 of the 54 staff members in Nursing Home A; none yielded salmonella. Two months later, another case was detected in Nursing Home A. Two cases were also detected in March and September 1997 in a second nursing home, Nursing Home B, located in an adjacent county. In November 1998, 18 months after the last patient with culture-confirmed salmonella infection was discharged from Nursing Home A, another case was detected there. Isolates were tested for antimicrobial susceptibility and found to be resistant to fluoroquinolones, and the Centers for Disease Control and Prevention (CDC) was invited to join the investigation. In November 1999, another patient with culture-confirmed salmonella infection was identified in Nursing Home A and found to reside in the room that had been occupied by the index patient, who had been discharged 29 months earlier. Thirty samples were collected on swabs from various surfaces and material in the room; two yielded S. enterica serotype Schwarzengrund. One of the two samples came from a foam mattress, which was then destroyed. In addition, cultures of urine and stool from the asymptomatic roommate of the newly infected patient yielded S. enterica serotype Schwarzengrund. In March and April of 2000, two more cases were identified. No additional cases occurred during 11 months of follow-up. Epidemiologic Investigation Information about the patients was obtained through reviews of medical charts and death certificates and included demographic data, the length of stay, use of antimicrobial agents, gastrointestinal symptoms, results of laboratory cultures, use of invasive procedures, and outcome. To identify possible mechanisms of transmission in Nursing Homes A and B, we interviewed staff members and conducted on-site inspections. To ascertain compliance with infection-control practices, we reviewed the annual surveys and inspection reports for the years 1994 through 1997, which we obtained from the Senior and Disabled Services Division of the Oregon Department of Human Services. To determine the risk factors for the transmission of fluoroquinolone-resistant salmonella infections in Nursing Home A, the facility with the greater number of patients, we conducted a case control study. Patients were defined as residents of Nursing Home A from whom fluoroquinolone-resistant S. enterica serotype Schwarzengrund was isolated between February 1996 and December Controls consisted of all 13 residents in the skilled-nursing care area of Nursing Home A who had negative cultures for salmonella in March To estimate the quantities of antimicrobial agents used at Nursing Home A as compared with other facilities, we conducted a focused survey of the use of antimicrobial agents in nursing homes in Oregon. We contacted the company that supplies pharmaceutical products to Nursing Home A, the second-largest supplier of pharmaceuticals to nursing homes in Oregon. We requested information on the number of prescriptions for antimicrobial agents that were filled between September 1 and December 16, 1998, at Nursing Home A as well as at three other nursing homes (not including Nursing Home B) of similar size and level of care. These other facilities were selected as a convenience sample by the pharmaceutical distribution company. To standardize the results for Patients in Nursing Home B Admission to and stay in nursing home Transfer to another nursing home or other institution Positive culture Discharge home 11 Roommates Overlapping stay 6 at Hospital X 5 4 Death 3 Death 2 Death 1 Patients in Nursing Home A Roommates Figure 1. Time Line of Events for 11 Patients with Positive Cultures for Salmonella enterica Serotype Schwarzengrund in Oregon, 1995 through Patient 6 resided in the room that had been occupied by Patient 1, the index patient. The numbers below the positiveculture symbols for Patients 5, 6, 7, and 10 indicate the number of positive cultures in a given month. N Engl J Med, Vol. 344, No. 21 May 24,

3 pharmaceutical use in different facilities, we calculated the rate of prescriptions for antimicrobial agents per 100 beds. Laboratory Investigation Isolates of S. enterica serotype Schwarzengrund were sent to the CDC for antimicrobial-susceptibility testing to determine the minimal inhibitory concentrations (MICs) of a panel of 16 antimicrobial agents by the broth-microdilution method with the use of the semiautomated Sensititre system (Trek Diagnostics, Westlake, Ohio). The guidelines of the National Committee for Clinical Laboratory Standards were used when they were available; an MIC of apramycin of at least 8 µg per milliliter and an MIC of ceftiofur of at least 32 µg per milliliter or more were considered indicative of a resistant organism. Isolates were also tested for resistance to ciprofloxacin with the use of the E test according to the manufacturer s recommendations (AB Biodisk, Solona, Sweden). To determine the mechanism of resistance, we used the polymerase chain reaction (PCR) to amplify the region of the gyra gene in the isolates that determines quinolone resistance and sequenced the region with the use of primers P1 and P2BIO. 13 Primers for PCR and sequencing were synthesized at the Biotechnology Core Facility of the CDC. PCR products were sequenced on an automated sequencer (model 377, Applied Biosystems Division, Perkin Elmer, Norwalk, Conn.). Isolates were also characterized by pulsed-field gel electrophoresis, as previously described. 14 Statistical Analysis We used the Wilcoxon rank-sum test to compare group means for nonparametric data. In the case control study, we calculated odds ratios and exact 95 percent confidence intervals, and we determined two-tailed P values with use of Fisher s exact test. All statistical analyses were performed with the use of Epi Info software (version 6.04, CDC, Stone Mountain, Ga.). Epidemiologic Information RESULTS A total of 11 persons with culture-confirmed fluoroquinolone-resistant S. enterica serotype Schwarzengrund infection (7 women and 4 men) were identified between February 1996 and April The median age of the patients was 85 years (range, 64 to 90). In nine patients, S. enterica serotype Schwarzengrund was first isolated from urine cultures; in one, it was first isolated from a stool culture, and in one from a wound culture. The urine samples had been obtained because the patients had a variety of clinical indications, including cloudy urine, fever, and abdominal pain. In five patients in whom the organism was first isolated from urine or wound culture, subsequent cultures of stool samples yielded S. enterica serotype Schwarzengrund. All patients had chronic medical problems. Five had indwelling devices (e.g., a central venous catheter, a urinary catheter, or a nasogastric tube). Of the 10 patients for whom we had complete information, 9 had been treated with fluoroquinolones in the six months before the collection of the specimen that yielded S. enterica serotype Schwarzengrund. The median interval between the use of a fluoroquinolone and a positive culture was 42 days (range, 7 to 124). Two patients were given fluoroquinolones as a treatment for the salmonella infection. Among the nine who were treated with fluoroquinolones before the first positive culture, the median duration of treatment was 7 days (range, 5 to 10). Four patients had gastrointestinal symptoms after beginning a course of fluoroquinolone therapy and before having a positive culture. Three patients had diarrhea, two had bloody stools, four had abdominal cramps, three had nausea, two had vomiting, and three had fever. Nine of the patients with S. enterica serotype Schwarzengrund infection were identified in Nursing Home A. An average of 100 patients reside in the facility, where the services provided include rehabilitation, long-term care, and skilled-nursing care. All nine patients resided in the skilled-nursing care area, which has a capacity of 21 beds. State inspectors identified no deficiencies in the infection-control procedures at Nursing Home A between 1994 and Two of the 30 environmental samples collected in December 1999 were positive for S. enterica serotype Schwarzengrund. These two samples were taken from a foam mattress in the index patient s room and the door handle of that room 29 months after the patient had been discharged. The mattress was stored under the patient s bed, was handled regularly by staff, and was used as a floor bed for the restraining of residents. The mattress was removed and destroyed in December During this outbreak, transmission apparently occurred in two hospitals (the hospital in the Philippines and Hospital X) and two nursing homes (Nursing Homes A and B). The index patient most likely acquired the infection during a hospital stay in the Philippines. On his return to Oregon, the patient was brought to Nursing Home A, where he shed the organism for at least 12 months; eight patients were subsequently infected at Nursing Home A. One of these patients was hospitalized at Hospital X on the same day as another patient who had had no contact with Nursing Home A. These two hospitalized patients were given a bath on the same day; however, the records do not indicate where the baths occurred or by whom they were given. The second patient apparently became infected at Hospital X and then went to Nursing Home B, where the infection spread to a roommate. Five residents of Nursing Home A met the definition of a patient with salmonella infection and were included in the case control study. They were compared with 13 residents of Nursing Home A who were also cared for by skilled nurses but whose stool samples did not yield S. enterica serotype Schwarzengrund in March The patients with salmonella infections did not differ significantly from the controls with respect to sex, race, age, or presence or absence of an indwelling device (Table 1). The median length of stay in the nursing home was longer for the patients with salmonella infection than for the controls (71 vs. 38 days, P=0.07). The patients with salmo N Engl J Med, Vol. 344, No. 21 May 24,

4 A NOSOCOMIAL OUTBREAK OF FLUOROQUINOLONE-RESISTANT SALMONELLA INFECTION TABLE 1. CHARACTERISTICS OF PATIENTS WITH SALMONELLA INFECTION AND CONTROLS IN NURSING HOME A IN OREGON, DECEMBER CHARACTERISTIC PATIENTS (N=5) CONTROLS (N=13) ODDS RATIO FOR SALMONELLA INFECTION (95% CI)* P VALUE Length of stay days Median Range Age yr Mean Range Female sex no. (%) 4 (80) 7 (54) 3.43 ( ) 0.60 Non-Hispanic white race no. (%) 5 (100) 12 (92) Undefined 1.00 Indwelling device no. (%) 3 (60) 5 (38) 2.40 ( ) 0.61 Treatment with antimicrobial agent no. (%) Fluoroquinolone 4 (80) 2 (15) ( ) 0.02 Trimethoprim sulfamethoxazole 2 (40) 1 (8) 8.00 ( ) 0.17 Other 3 (60) 5 (38) 2.40 ( ) 0.61 Death no. (%) 3 (60) 5 (38) 2.40 ( ) 0.61 *CI denotes confidence interval. Because none of the patients were of another racial or ethnic group, an odds ratio could not be computed but approaches infinity. Reported data are for use during the six months before a culture was obtained nella infection were more likely than the controls to have taken fluoroquinolones in the six months before a culture was obtained (4 of the 5 patients, as compared with 2 of the 13 controls; odds ratio, 22.0; 95 percent confidence interval, 1.06 to 1177; P=0.02) (Table 1). These four patients took fluoroquinolones a median of 14 days (range, 7 to 55) before their first positive culture. Use of trimethoprim sulfamethoxazole or other antimicrobial agents in the six months before a culture was obtained was not associated with the acquisition of salmonella infection. Three of the 5 patients died, as compared with 5 of the 13 controls (P=0.61). In the three patients with salmonella infection who died, the intervals between the last positive culture and death were 3 days, 23 days, and 7 months. The causes of death were pneumonia, intracerebral hemorrhage, and myeloid dysplasia, respectively. Nursing Home A used more fluoroquinolones than the other nursing homes we studied; fluoroquinolones were the most commonly used antimicrobial agents in these facilities (Fig. 2). Two patients with S. enterica serotype Schwarzengrund infection were identified in Nursing Home B. An average of 100 patients reside in this facility, where the services provided include long-term and subacute care. The two patients resided in the subacute care area, and for two months they shared the same room. State inspectors had previously noted deficiencies in infection-control practices at the nursing home. Most recently, the facility had been cited for the failure to add a cleanser during the washing of soiled linens. Laboratory Results The isolates from the outbreak were all resistant to ampicillin, ciprofloxacin, and trimethoprim sulfamethoxazole, and some were resistant to a variety of other agents (Table 2). All isolates from the outbreak had similar patterns on pulsed-field gel electrophoresis, which were also similar to the only previous isolate of fluoroquinolone-resistant salmonella in the United States, a sample of S. enterica serotype Schwarzengrund isolated in New York in 1995 (Fig. 3). 11 Four control isolates of S. enterica serotype Schwarzengrund collected in Oregon between 1995 and 1998 but not associated with the outbreak had different patterns on pulsed-field gel electrophoresis and were sensitive to ciprofloxacin. The nine ciprofloxacin-resistant isolates we tested had the same two point mutations in the region of gyra that determines quinolone resistance, whereas the control isolates had the wild-type sequence (Table 3). The two point mutations were identical to those seen in the 1995 isolate of ciprofloxacin-resistant S. enterica serotype Schwarzengrund from New York. DISCUSSION This report describes the first outbreak of fluoroquinolone-resistant salmonella infection in the United States. Between February 1996 and April 2000, 11 patients with laboratory-confirmed infection with fluoroquinolone-resistant S. enterica serotype Schwarzengrund were identified in Portland, Oregon. Transmission probably occurred in two nursing homes and N Engl J Med, Vol. 344, No. 21 May 24,

5 No. of PrescriptionsŁ per 100 Beds Fluoroquinolone Trimethoprim sulfamethoxazole Penicillin Cephalosporin Amoxicillin clavulanic acid Macrolide antibiotic Nitrofurantoin 0 Nursing Home A Nursing Home 1 Nursing Home 2 Nursing Home 3 Figure 2. Number of Prescriptions for Antimicrobial Agents in Four Nursing Homes in Oregon, September 1, 1998, to December 16, Nursing Home A was the only one of the nursing homes that had patients with culture-confirmed infections with Salmonella enterica serotype Schwarzengrund. TABLE 2. RANGE OF CONCENTRATIONS TESTED, MINIMAL INHIBITORY CONCENTRATIONS OF 16 ANTIMICROBIAL AGENTS, AND NUMBER OF THE ISOLATES OF SALMONELLA ENTERICA SEROTYPE SCHWARZENGRUND THAT WERE RESISTANT. ANTIMICROBIAL AGENT RANGE OF CONCENTRATIONS EVALUATED MIC* ISOLATES WITH RESISTANCE µg/ml no. (%) Amikacin 4 32 «4 0 Amoxicillin clavulanic / /4 0 acid Ampicillin 2 64 >32 9 (100) Apramycin 2 16 >32 7 (78) Ceftiofur Ceftriaxone « Cephalothin (11) Chloramphenicol 4 32 >32 5 (56) Ciprofloxacin >4 9 (100) Gentamicin (33) Kanamycin «16 0 Nalidixic acid 4 64 >256 9 (100) Streptomycin «32 0 Sulfamethoxazole >512 9 (100) Tetracycline 4 64 >32 8 (89) Trimethoprim sulfamethoxazole / >4/>76 9 (100) *MIC denotes minimal inhibitory concentration. This agent is for veterinary use only. The result of the E test for ciprofloxacin on the isolate from the index patient was 4 µg per milliliter, and the results from all other isolates were more than 32 µg per milliliter. one hospital in the Portland area. The index patient apparently acquired the infection while hospitalized in the Philippines in 1994 and was subsequently transferred to one of the nursing homes. Of approximately 8000 salmonella isolates tested at the CDC between 1994 and 1998, only 1 other isolate was resistant to fluoroquinolone. 11 That isolate was also S. enterica serotype Schwarzengrund and was obtained from a person who had also been hospitalized in the Philippines in As in an earlier study that examined previous use of antimicrobial agents as a risk factor for drug-resistant salmonellosis, 15 our case control study demonstrated that fluoroquinolone use in the six months before the date of specimen collection was significantly associated with an increased likelihood of a fluoroquinolone-resistant salmonella infection. Although trimethoprim sulfamethoxazole was also widely used in Nursing Home A, its use was not associated with illness in the case control study. We used a wide window of exposure because we were unable to ascertain precisely the time of onset of illness through review of the medical charts, and the onset was often not associated with specific reported symptoms. The longer exposure period probably increased the extent of nondifferential misclassification (i.e., the incorrect classification of exposure for an equal proportion of patients and controls), making it more difficult to demonstrate an association. Our finding suggests that the use of fluoroquinolones may select for and contribute to the transmission of this fluoroquinolone-resistant strain of S. enterica serotype Schwarzengrund, perhaps by lowering the inoculum required to cause infection. 7 The isolates from the outbreak in Oregon and the isolate from the patient in New York had similar patterns on pulsed-field gel electrophoresis, suggesting 1576 N Engl J Med, Vol. 344, No. 21 May 24,

6 A NOSOCOMIAL OUTBREAK OF FLUOROQUINOLONE-RESISTANT SALMONELLA INFECTION Control Control Environmental swab New York patient Control Figure 3. Results of Pulsed-Field Gel Electrophoresis of Fluoroquinolone-Resistant Isolates of Salmonella enterica Serotype Schwarzengrund from Oregon and New York. Lanes 1, 8, and 14 show S. enterica serotype Newport controls; lanes 2 through 7 and 9 through 11 show isolates from the nine patients in Oregon; lane 12 shows the environmental isolate from a foam mattress; and lane 13 shows the isolate obtained in 1995 from the patient in New York. that the infections arose from a common source. The sequencing results support this hypothesis: all isolates had the same two mutations in the gyra gene. Variability in the mutations of the gyra gene has been demonstrated among other fluoroquinolone-resistant isolates. 16,17 Furthermore, these laboratory data are consistent with the epidemiologic data; both the New York patient and the index patient in the Oregon outbreak had been hospitalized in the Philippines. Further investigation revealed that these two patients had been hospitalized in the same facility in Manila, the Philippines. We were also able to ascertain that in 1992 and 1993, the prevalence of fluoroquinolone resistance in nontyphoidal salmonella in the Philippines was 2.5 percent, 18 and by 1998 it had increased to 4.7 percent (Turnidge J: personal communication). In the Philippines, several factors may contribute to increasing fluoroquinolone resistance, including overthe-counter availability of fluoroquinolones 18 and the use of fluoroquinolones in food animals. 19 The variability in resistance to other antimicrobial agents suggests that different isolates may have acquired different determinants of resistance (e.g., plasmids and integrons). This investigation documents the transmission of fluoroquinolone-resistant salmonella in the United States. Because fluoroquinolones are the drug of choice for the treatment of severe salmonella infections in adults, 6 the potential continued emergence and dissemination of fluoroquinolone-resistant salmonella are of great concern. In 1999, the Florida Department of Health reported an outbreak of nosoco- N Engl J Med, Vol. 344, No. 21 May 24,

7 TABLE 3. SEQUENCES OF THE FLUOROQUINOLONE-RESISTANCE DETERMINING REGIONS OF THE gyra GENE AND THE MINIMAL INHIBITORY CONCENTRATION OF CIPROFLOXACIN IN FLUOROQUINOLONE-SUSCEPTIBLE AND FLUOROQUINOLONE-RESISTANT ISOLATES. SOURCE OF ISOLATE SALMONELLA SEROTYPE gyra CODON MIC OF CIPROFLOXACIN (µg/ml)* 80 HIS 81 GLY 82 ASP 83 SER 84 ALA 85 VAL 86 TYR 87 ASP 88 THR GenBank Wild type CAC GGC GAT TCC GCA GTG TAT GAC ACC Oregon but not related Schwarzengrund «0.01 to outbreak Outbreak in Oregon Schwarzengrund _T_ _G_»4 New York Schwarzengrund _T_ _G_»4 *MIC denotes minimal inhibitory concentration. An MIC of 4 µg per milliliter or higher is considered to indicate resistance. The GenBank accession number is X The change results in the substitution of phenylalanine for serine at position 83. The change results in the substitution of glycine for aspartic acid at position 87. mially acquired fluoroquinolone-resistant S. enterica serotype Senftenberg infections at a medical facility; the similarities between the Florida and Oregon outbreaks are striking. 20 The transmission, perhaps through environmental contamination, apparently occurred within medical facilities a circumstance that should raise questions about the transmissibility of a highly resistant organism. The reservoir for most nontyphoidal salmonella in the United States is food-producing animals, and therefore the emerging resistance of salmonella is largely a consequence of the use of antimicrobial agents in animals. This outbreak was unusual because it occurred in medical facilities and involved personto-person transmission. In the developing world, nosocomial and community-acquired multidrug-resistant salmonellosis has been a recurrent problem. 4,5,21 The origins of resistance in the developing world are unknown. Since enhanced infection-control practices were implemented in the United States in the 1970s, nosocomial outbreaks of salmonella have been rare. The emergence of highly resistant salmonella in an environment characterized by heavy use of antimicrobial agents may provide a setting that is conducive to nosocomial transmission. 4 The outbreak we studied was difficult to control and lingered for several years. The environmental contamination and the duration of fecal shedding, as well as the characteristics of the facilities, 22 may have contributed to the prolonged nature of the outbreak. Those involved in infection control in hospitals, as well as public health authorities, should be aware that fluoroquinolone-resistant salmonella has arrived in the United States, and clinicians should consider drugresistant salmonella in the differential diagnosis of nosocomial infections. Nursing homes should evaluate their infection-control procedures and implement appropriate changes. Additional studies may be needed to determine where antimicrobial agents are being used in a nonjudicious manner. Finally, this outbreak underscores the need for the judicious use of antimicrobial agents in all domestic and international settings. REFERENCES 1. Mead PS, Slutsker L, Dietz V, et al. Food-related illness and death in the United States. Emerg Infect Dis 1999;5: Tauxe RV. Salmonella: a postmodern pathogen. J Food Protect 1991; 54: Angulo FJ, Johnson KR, Tauxe RV, Cohen ML. Origins and consequences of antimicrobial-resistant nontyphoidal Salmonella: implications for the use of fluoroquinolones in food animals. Microb Drug Resist 2000; 6: Riley LW, Ceballos BS, Trabulsi LR, Fernandes de Toledo MR, Blake PA. The significance of hospitals as reservoirs for endemic multiresistant Salmonella typhimurium causing infection in urban Brazilian children. J Infect Dis 1984;150: Kumar A, Nath G, Bhatia BD, Bhargava V, Loiwal V. An outbreak of multidrug resistant Salmonella typhimurium in a nursery. Indian Pediatr 1995;32: Miller SI, Pegues DA. Salmonella species, including Salmonella typhi. In: Mandell GL, Bennett JE, Dolin R, eds. Mandell, Douglas, and Bennett s principles and practices of infectious diseases. 5th ed. Vol. 2. Philadelphia: Churchill Livingstone, 2000: Cohen ML, Tauxe RV. Drug-resistant Salmonella in the United States: an epidemiologic perspective. Science 1986;234: Lee LA, Puhr ND, Maloney EK, Bean NH, Tauxe RV. Increase in antimicrobial-resistant Salmonella infections in the United States, J Infect Dis 1994;170: Piddock LJ, Griggs DJ, Hall MC, Jin YF. Ciprofloxacin resistance in clinical isolates of Salmonella typhimurium obtained from two patients. Antimicrob Agents Chemother 1993;37: Pers C, Sogaard P, Pallesen L. Selection of multiple resistance in Salmonella enteritidis during treatment with ciprofloxacin. Scand J Infect Dis 1996;28: Herikstad H, Hayes P, Mokhtar M, Fracaro ML, Threlfall EJ, Angulo FJ. Emerging quinolone-resistant Salmonella in the United States. Emerg Infect Dis 1997;3: Vasallo FJ, Martin-Rabadan P, Alcala L, Garcia-Lechuz JM, Rodriguez-Creixems M, Bouza E. Failure of ciprofloxacin therapy for invasive nontyphoidal salmonellosis. Clin Infect Dis 1998;26: N Engl J Med, Vol. 344, No. 21 May 24,

8 A NOSOCOMIAL OUTBREAK OF FLUOROQUINOLONE-RESISTANT SALMONELLA INFECTION 13. Griggs DJ, Gensberg K, Piddock LJ. Mutations in gyra gene of quinolone-resistant Salmonella serotypes isolated from humans and animals. Antimicrob Agents Chemother 1996;40: Standardized molecular subtyping of foodborne bacterial pathogens by pulsed-field gel electrophoresis. Atlanta: Centers for Disease Control and Prevention, Pavia AT, Shipman LD, Wells JG, et al. Epidemiologic evidence that prior antimicrobial exposure decreases resistance to infection by antimicrobial-sensitive Salmonella. J Infect Dis 1990;161: Piddock LJ. Mechanisms of resistance to fluoroquinolones: state-ofthe-art Drugs 1995;49:Suppl 2: Acar JF, Goldstein F W. Trends in bacterial resistance to fluoroquinolones. Clin Infect Dis 1997;24:Suppl 1:S67-S Turnidge J. Epidemiology of quinolone resistance: Eastern hemisphere. Drugs 1995;49:Suppl 2: Use of quinolones in food animals and potential impact on human health. Geneva: World Health Organization, June (WHO/EMC/ ZDI/98.12.) 20. Olsen S, Blackmore C, DeBess E, et al. Emergence of fluoroquinolone-resistant Salmonella infections in the United States: nosocomial outbreaks suggest a changing epidemiology. In: Abstracts of the International Conference on Emerging Infectious Diseases, Atlanta, July 16 19, Washington, D.C.: American Society for Microbiology, 2000:144. abstract. 21. Lepage P, Bogaerts J, Van Goethem C, et al. Community-acquired bacteraemia in African children. Lancet 1987;1: Li J, Birkhead GS, Strogatz DS, Coles FB. Impact of institution size, staffing patterns, and infection control practices on communicable disease outbreaks in New York State nursing homes. Am J Epidemiol 1996;143: Copyright 2001 Massachusetts Medical Society. ELECTRONIC ACCESS TO THE JOURNAL S CUMULATIVE INDEX At the Journal s site on the World Wide Web ( you can search an index of all articles published since January You can search by author, subject, title, type of article, or date. The results will include the citations for the articles plus links to the abstracts of articles published since Single articles and past issues of the Journal can also be ordered for a fee through the Internet ( N Engl J Med, Vol. 344, No. 21 May 24,